Turn-on and ratiometric mercury sensing in water with a red-emitting probe

Article abstract of DOI:10.1021/ja068879r

The synthesis, photophysical properties, and Hg(II) binding of a red-emitting sensor for mercuric ion are presented. 2-{11-[(2-{[Bis-(2- ethylsulfanylethyl)amino]methyl}phenylamino)methyl]-3-hydroxy-10-oxo-10H- benzo[c]xanthen-7-yl{benzoic acid (MS5) is b

Full text of DOI:10.1021/ja068879r

Published on Web 04/13/2007
Turn-On and Ratiometric Mercury Sensing in Water with a
Red-Emitting Probe
Elizabeth M. Nolan and Stephen J. Lippard*
Contribution from the Department of Chemistry, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139
Received December 12, 2006; E-mail: lippard@mit.edu
Abstract: The synthesis, photophysical properties, and Hg(II) binding of a red-emitting sensor for mercuric
ion are presented. 2-{11-[(2-{[Bis-(2-ethylsulfanylethyl)amino]methyl}phenylamino)methyl]-3-hydroxy-10-
oxo-10H-benzo[c]xanthen-7-yl}benzoic acid (MS5) is based on the seminaphthofluorescein chromophore
and employs a thioether-rich metal-binding unit. This sensor affords both turn-on and single-excitation dual-
emission ratiometric Hg(II) detection in aqueous solution. The fluorescence response of MS5 is Hg(II)-
specific, and the probe is selective for Hg(II) over alkali and alkaline earth metals, most divalent first-row
transition metal ions, and the Group 12 congeners Zn(II) and Cd(II). MS5 binds Hg(II) reversibly and can
be recycled. The EC50 for 1 µM MS5 is 910 nM, and a lower detection limit of 50 nM is obtained when
employing 500 nM probe. X-ray crystallographic studies using a salicylaldehyde-based model of MS5 are
also presented. 2-[(2-{[Bis-(2-ethylsulfanylethyl)amine]methyl}phenylamine)methyl]phenol coordinates
Hg(II) with two thioether sulfur atoms, two amino nitrogen atoms, and a phenol oxygen atom arranged in
a distorted trigonal bipyramidal geometry. Studies of natural water samples spiked with mercuric salts indicate
that MS5 can rapidly detect Hg(II) in such complex solutions and demonstrate its potential utility in the
field.
Introduction
fluorescent small-molecule Hg(II) sensors presented to date are
quenched upon Hg(II) coordination, rely on an irreversible
A variety of natural and anthropogenic environmental con-
taminants pose serious problems for human health and ecology.
Mercury pollution is a topic of recent concern1-6 and has
sparked interest in the design of new chemical tools and tactics
for its detection. Small-molecule sensors that provide selective
and immediate response to Hg(II), observable visually or with
simple spectroscopic instrumentation, are particularly valuable.
Such probes can ultimately be employed in basic laboratory
assays, adapted for measurements in the field through their
incorporation into portable fiber optic devices, and, as is the
case for Pb(II) detection, fabricated into commercial indictors
for household use.
Hg(II)-dependent chemical reaction to give fluorescence turn-
on, and/or require organic solvent systems, features that may
be problematic depending on a given application. Achieving
(14) Yoon, J.; Ohler, N. E.; Vance, D. H.; Aumiller, W. D.; Czarnik, A. W.
Tetrahedron Lett. 1997, 38, 3845-3848.
(15) Hennrich, G.; Sonnenschein, H.; Resch-Genger, U. J. Am. Chem. Soc. 1999,
121, 5073-5074.
(16) Prodi, L.; Bargossi, C.; Montalti, M.; Zaccheroni, N.; Su, N.; Bradshaw, J.
S.; Izatt, R. M.; Savage, P. B. J. Am. Chem. Soc. 2000, 122, 6769-
6770.
(17) Rurack, K.; Kollmannsberger, M.; Resch-Genger, U.; Daub, J. J. Am. Chem.
Soc. 2000, 122, 968-969.
(18) Hennrich, G.; Walther, W.; Resch-Genger, U.; Sonnenschein, H. Inorg.
Chem. 2001, 40, 641-644.
(19) Sakamoto, H.; Ishikawa, J.; Nakao, S.; Wada, H. Chem. Commun. 2000,
2395-2396.
Fluorescence is a powerful optical technique for detecting
low concentrations of analytes. A number of fluorescence-based
Hg(II) probes have been reported and include polymers,7,8
foldamers,9 biomolecules,10-12 and small molecules.13-32 Many
(20) Guo, X.; Qian, X.; Jia, L. J. Am. Chem. Soc. 2004, 126, 2272-2273.
(21) Me´tivier, R.; Leray, I.; Valeur, B. Chem. Eur. J. 2004, 10, 4480-4490.
(22) Dickerson, T. J.; Reed, N. N.; LaClair, J. J.; Janda, K. D. J. Am. Chem.
Soc. 2004, 126, 16582-16586.
(23) Zhang, G.; Zhang, D.; Yin, S.; Yang, X.; Shuai, Z.; Zhu, D. Chem. Commun.
2005, 2161-2163.
(24) Liu, B.; Tian, H. Chem. Commun. 2005, 3156-2158.
(25) Caballero, A.; Mart´ınez, R.; Lloveras, V.; Ratera, I.; Vidal-Gancedo, J.;
Wurst, K.; Ta´rraga, A.; Molina, P.; Veciana, J. J. Am. Chem. Soc. 2005,
127, 15666-15667.
(1) Renzoni, A.; Zino, F.; Franchi, E. EnViron. Res. Sect. A 1998, 77, 68-72.
(2) Nendza, M.; Herbst, T.; Kussatz, C.; Gies, A. Chemosphere 1997, 35,
1875-1885.
(3) Bolger, P. M.; Schwetz, B. A. New Engl. J. Med. 2002, 347, 1735-1736.
(4) Harris, H. H.; Pickering, I. J.; George, G. N. Science 2003, 301, 1203.
(5) Boening, D. W. Chemosphere 2000, 40, 1335-1351.
(26) Ros-Lis, J. V.; Marcos, M. D.; Ma´rtinez-Ma´n˜ez, R.; Rurack, K.; Soto, J.
Angew. Chem., Int. Ed. 2005, 44, 4405-4407.
(27) Chen, Q.-Y.; Chen, C.-F. Tetrahedron Lett. 2005, 46, 165-168.
(28) Moon, S.-Y.; Youn, N. J.; Park, S. M.; Chang, S.-K. J. Org. Chem. 2005,
70, 2394-2397.
(6) Clarkson, T. W.; Magos, L.; Myers, G. J. New Engl. J. Med. 2003, 349,
1731-1737.
(7) Fan, L.-J.; Zhang, Y.; Jones, W. E. Macromolecules 2005, 38, 2844-2849.
(8) Kim, I.-B.; Bunz, U. H. F. J. Am. Chem. Soc. 2006, 128, 2818-2819.
(9) Zhao, Y.; Zhong, Z. J. Am. Chem. Soc. 2006, 128, 9988-9989.
(10) Cheng, P.; He, C. J. Am. Chem. Soc. 2004, 126, 728-729.
(11) Ono, A.; Togashi, H. Angew. Chem., Int. Ed. 2004, 43, 4300-4302.
(12) Matsushita, M.; Meijler, M. M.; Wirsching, P.; Lerner, R. A.; Janda, K.
D. Org. Lett. 2005, 7, 4943-4946.
(29) Yu, Y.; Lin, L.-R.; Yang, K.-B.; Zhong, X.; Huang, R.-B.; Zheng, L.-S.
Talanta 2006, 69, 103-106.
(30) Kim, S. H.; Kim, J. S.; Park, S. M.; Chang, S.-K. Org. Lett. 2006, 8, 371-
374.
(31) Kim, S. H.; Song, K. C.; Ahn, S.; Kang, Y. S.; Chang, S.-K. Tetrahedron
Lett. 2006, 47, 497-500.
(32) Song, K. C.; Kim, J. S.; Park, S. M.; Chung, K.-C.; Ahn, S.; Chang, S.-K.
Org. Lett. 2006, 8, 3413-3416.
(13) Chae, M.-Y.; Czarnik, A. W. J. Am. Chem. Soc. 1992, 114, 9704-9705.
9
5910
J. AM. CHEM. SOC. 2007, 129, 5910-5918
10.1021/ja068879r CCC: $37.00 © 2007 American Chemical Society

Mercury Sensing in Water with a Red-Emitting Probe
A R T I C L E S
Figure 1. Structures of metal ion sensors and seminaphthofluorescein 3.
water compatibility is important33 and often a challenge in
Hg(II) sensor design. Recently, significant efforts have been
made to prepare Hg(II)-sensitive molecules that operate in purely
aqueous solution and afford fluorescence enhancement and/or
a ratiometric response.34-44 Many of these detectors utilize
fluorescein,35,43 rhodamine,37,40,44 BODIPY,42 or naphthalimide38
reporting groups and exhibit emission in the visible region of
the spectrum, with λmax e 560 nm. Examples of Hg(II) sensors
with lower energy emission are limited. A phenoxazinone
appended with a thioether-rich macrocycle affords maximum
emission at ca. 600 nm in aqueous solution, but Hg(II)
coordination results in fluorescence quenching.45 MS4 (Figure
1) exhibits emission enhancement centered at 624 nm following
Hg(II) binding, but the change is small (∼2-fold) at neutral pH.35
Two probes that respond to Hg(II) and give near-infrared
emission in organic solvents have been reported, but their
applicability to aqueous milieu was not documented.46,47
Designing improved Hg(II) sensors that emit at relatively long
wavelengths, exhibit Hg(II) sensitivity and selectivity, and have
a robust and immediate response to Hg(II) in water is an
important goal. Such probes will be useful in a number of
applications, including multicolor analyses requiring sensors that
emit at different wavelengths.
In the present article, we address this need and present the
synthesis and characterization of Mercury Sensor 5 (MS5,
Scheme 1). This sensor is based on a seminaphthofluorescein
chromophore, which is water-compatible and affords red-shifted
emission relative to fluorescein, rhodamine, and other commonly
used reporters. Given the high affinity of Hg(II) for soft sulfur
donors, we chose to incorporate the thioether-rich metal-binding
unit previously designed for fluorescein-based MS1 (Figure 1).
Because of these design features, MS5 has a number of
advantages that include (i) turn-on and ratiometric Hg(II)
detection, (ii) rapid and Hg(II)-specific fluorescence response,
(iii) reversibility, and (iv) a detection limit for Hg(II) in the
nanomolar range. In addition, MS5 exhibits selectivity for Hg(II)
over many competing species, including those found in natural
water samples.
Experimental Section
(33) Czarnik, A. W. Acc. Chem. Res. 1994, 27, 302-308.
(34) Nolan, E. M.; Lippard, S. J. J. Am. Chem. Soc. 2003, 125, 14270-14271.
(35) Nolan, E. M.; Lippard, S. J. J. Mater. Chem. 2005, 27-28, 2778-2783.
(36) Mello, J. V.; Finney, N. S. J. Am. Chem. Soc. 2005, 127, 10124-10125.
(37) Yang, Y.-K.; Yook, K.-J.; Tae, J. J. Am. Chem. Soc. 2005, 127, 16760-
16761.
Reagents. Anhydrous methanol, chloroform, and 1,2-dichloroethane
(DCE) were purchased from Aldrich and used as received. Sodium
triacetoxyborohydride and mercury(II) perchlorate hydrate were also
obtained from Aldrich. N′,N′,N′′,N′′-tetrakis(2-picolyl)ethylenediamine
(TPEN) was purchased from Invitrogen. The aniline-derivatized ligand
134 and seminaphthofluorescein aldehyde 248 were synthesized according
to previously published procedures. A portion of 2-(10-hydroxy-3-
phenylaminomethyl-3H-benzo[c]xanthen-7-yl)benzoic acid (3), a syn-
thetic precursor to 2, was prepared as described in the literature48 and
further purified by preparative thin-layer chromatography (TLC) on
silica gel (20:1 CHCl3/MeOH) for use in spectroscopic studies.
CAUTION! The isolated Hg(II) complexes described below contain
perchlorate ion, which can detonate explosively and without warning.
(38) Wang, J.; Qian, X. Chem. Commun. 2006, 109-111.
(39) Yoon, S.; Albers, A. E.; Wong, A. P.; Chang, C. J. J. Am. Chem. Soc.
2005, 127, 16030-16031.
(40) Zheng, H.; Qian, Z.-H.; Xu, L.; Yuan, F.-F.; Lan, L.-D.; Xu, J.-G. Org.
Lett. 2006, 8, 859-861.
(41) Wang, J.; Qian, X.; Cui, J. J. Org. Chem. 2006, 71, 4308-4311.
(42) Wang, J.; Xuhong, Q. Org. Lett. 2006, 8, 3721-3724.
(43) Nolan, E. M.; Racine, M. E.; Lippard, S. J. Inorg. Chem. 2006, 45, 2742-
2749.
(44) Ko, S.-K.; Yang, W.-K.; Tae, J.; Shin, I. J. Am. Chem. Soc. 2006, 128,
14150-14155.
(45) Descalzo, A. B.; Mart´ınez-Ma´n˜ez, R.; Radeglia, R.; Rurack, K.; Soto, J. J.
Am. Chem. Soc. 2003, 125, 3418-3419.
(46) Rurack, K.; Resch-Genger, U.; Spieles, M.; Bricks, J. L. Chem. Commun.
2000, 2103-2104.
(47) Zhu, X.-J.; Fu, S.-T.; Wong, W.-K.; Guo, J.-P.; Yong, W.-Y. Angew. Chem.,
Int. Ed. 2006, 45, 3150-3154.
(48) Chang, C. J.; Jaworski, J.; Nolan, E. M.; Sheng, M.; Lippard, S. J. Proc.
Natl. Acad. Sci. U.S.A. 2004, 101, 1129-1134.
9
J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007 5911

A R T I C L E S
Nolan and Lippard
Scheme 1. Synthesis of MS5 and Its Salicylaldehyde-Based Analogue L
Although we encountered no problems with the isolated compounds,
all due precautions should be taken.
an orange oil. Chromatography on Al2O3 with a solvent gradient (2:1
hexanes/EtOAc to 100% EtOAc) afforded the product as an orange oil
that solidified upon cooling at 4 °C (812 mg, 65%): TLC Rf ) 0.60
(Al2O3, 2:1 hexanes/EtOAc); 1H NMR (CDCl3, 300 MHz) δ 1.17 (6H,
t), 2.37 (4H, q), 2.58 (4H, m), 2.65 (4H, m), 3.70 (2H, s), 4.48 (2H, s),
6.78-6.91 (4H, m), 7.05 (1H, dd), 7.16-7.23 (3H, m); 13C NMR
(CDCl3, 125 MHz) δ 14.79, 25.88, 29.34, 47.65, 53.02, 58.35, 113.18,
116.33, 119.25, 119.75, 123.64, 124.72, 128.57, 128.90, 130.29, 147.42,
157.16; FTIR (NaCl disk, cm-1) 3217 (s, br), 3043 (s), 2963 (s), 2924
(s), 2869 (s), 2845 (s), 2817 (s), 2717 (w), 1606 (s), 1588 (s), 1508
(s), 1490 (s), 1455 (s), 1375 (m), 1314 (m), 1294 (m), 1250 (m), 1149
(w), 1119 (m), 1102 (m), 1050 (m), 1036 (m), 983 (m), 932 (m), 855
(m), 839 (m), 753 (s), 723 (m), 659 (w), 629 (w), 565 (w), 531 (w);
HRMS (ESI) calcd, 405.2029 ([M + H]+), found 405.2019.
[Hg(L)](ClO4)2‚CH3OH (6). A portion (9.7 mg, 0.024 mmol) of
Hg(ClO4)2‚H2O was suspended in 0.5 mL of MeOH, to which a solution
of 5 (10 mg, 0.025 mmol) in 0.5 mL of MeOH was added dropwise as
the solution turned yellow. The solution was sonicated for 5 min and
filtered. Colorless, thin needle crystals suitable for X-ray crystal-
lographic analysis formed over the course of 48 h by vapor diffusion
of Et2O into the reaction at room temperature. The solvents were
decanted, and the crystals were washed with Et2O (4 × 3 mL) and
dried in vacuo (11.5 mg, 48%): FTIR (KBr, cm-1) 3546 (m, br), 3192
(m, br), 3069 (w), 2969 (w), 2932 (w), 2865 (w), 2835 (w), 2700 (w),
2605 (w), 2552 (w), 1609 (m), 1594 (m), 1496 (m), 1460 (s), 1421
(m), 1384 (m), 1368 (m), 1300 (w), 1258 (m), 1218 (w), 1188 (w),
1102 (s, br), 1084 (s), 1055 (s), 1047 (s), 1004 (m), 932 (m), 922 (m),
910 (m), 884 (m), 870 (m), 853 (m), 832 (w), 801 (w), 770 (s), 729
(w), 717 (m), 672 (w), 622 (s), 575 (w), 543 (w), 521 (w), 494 (w),
484 (w), 468 (w); HRMS (ESI) calcd 605.1584 ([M -2(ClO4) - H]+),
found 605.1578. Anal. Calcd for C23H36Cl2N2O10S2Hg: C, 33.04; H,
4.34; N, 3.35. Found: C, 33.01; H, 4.44; N, 3.30.
Materials and Methods. Merck 254 silica gel 60 plates (0.25 mm
thickness) were used for analytical TLC and were viewed with UV
light. Whatman silica gel 60 plates (1 mm thickness) were used as the
solid phase for preparative TLC. NMR spectra were collected on a
Varian 300 or 500 MHz spectrophotometer, and the spectra were
referenced to internal standards. An Avatar FTIR instrument was used
to obtain IR spectra. High-resolution mass spectra were collected by
staff at the MIT Department of Chemistry Instrumentation Facility.
2-{11-[(2-{[Bis-(2-ethylsulfanylethyl)amino]methyl}phenylamino)-
methyl]-3-hydroxy-10-oxo-10H-benzo[c]xanthen-7-yl}benzoic Acid
(4, MS5). To 9 mL of a 7:2 CHCl3/MeOH mixture were added 1 (71
mg, 0.17 mmol) and 2 (52 mg, 0.17 mmol), which gave a red-brown
solution that was stirred at room temperature. After 24 h, 3 mL of
DCE and NaB(OAc)3H (40 mg, 0.19 mmol) were added, and the
reaction was stirred for an additional 24 h, during which time it became
wine-colored. The solvents were removed under reduced pressure, and
preparative TLC on silica gel (20:1 CHCl3/MeOH) yielded pure MS5
as a purple solid (38 mg, 32%): TLC Rf ) 0.64 (9:1 CHCl3/MeOH);
1
mp >300 °C dec; H NMR (CD3OD, 500 MHz) δ 0.94 (6H, t), 2.08
(4H, m), 2.24 (4H, m), 2.39 (2H, m), 2.47 (2H, m), 3.55 (2H, s), 4.70
(2H, q), 6.64 (1H, t), 6.71 (1H, d), 6.92 (1H, d), 6.98 (3H, m), 7.13
(1H, d), 7.20 (1H, d), 7.28 (2H, m), 7.34 (1H, m), 7.61 (2H, m), 8.10
(1H, d), 8.27 (1H, m); FTIR (KBr, cm-1) 3428 (w, br), 3052 (w), 2957
(w), 2921 (w), 2857 (w), 2800 (w), 1752 (w), 1641 (m), 1600 (s), 1585
(s), 1563 (m), 1504 (m), 1470 (s), 1443 (s), 1377 (m), 1339 (m), 1309
(m), 1251 (m), 1143 (m), 1091 (m), 1070 (w), 1046 (w), 1004 (w),
961 (w), 931 (w), 877 (w), 862 (w), 827 (w), 789 (w), 748 (w), 726
(w), 704 (w), 678 (w), 643 (w), 613 (w), 596 (w), 526 (w), 448 (w);
HRMS (ESI) calc, 691.2306 ([M - H]-), found 691.2325.
2-[(2-{[Bis-(2-ethylsulfanylethyl)amine]methyl}phenylamine)-
methyl]phenol (5, L). Salicylaldehyde (376 mg, 3.08 mmol) and N-(2-
aminobenzyl)-3,9-dithia-6-azaundecane (1, 914 mg, 3.07 mmol) were
combined in 40 mL of EtOAc and stirred for 10 h at room temperature.
The solvent was removed in vacuo to yield an orange-yellow oil. The
oil was dissolved in 30 mL of DCE, and NaB(OAc)3H (700 mg, 3.33
mmol) was added. The reaction was stirred overnight at room
temperature, diluted with 40 mL of dichloromethane, and washed with
saturated brine (3 × 80 mL) and then water (1 × 80 mL). The organic
portion was dried over MgSO4, and the solvent was removed to yield
[Hg(TPEN)](ClO4)2 (7). A portion (7.8 mg, 0.195 mmol) of
Hg(ClO4)2‚H2O in 1 mL of MeOH was added dropwise to a solution
of TPEN (7.9 mg, 0.184 mmol) in 1 mL of MeOH. The solution was
mixed with a pipet and filtered. Colorless rodlike crystals suitable for
X-ray crystallography formed over the course of 48 h from vapor
diffusion of Et2O into the reaction mixture at room temperature. The
solvents were decanted, and the crystals were washed with Et2O (4 ×
3 mL) and dried in vacuo (8.6 mg, 57%): FTIR (KBr, cm-1) 3107
(w), 3078 (w), 3033 (w), 2963 (w), 2937 (w), 2911 (w), 2895 (w),
9
5912 J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007

Mercury Sensing in Water with a Red-Emitting Probe
A R T I C L E S
2859 (w), 1600 (m), 1572 (w), 1486 (m), 1467 (m), 1455 (m), 1441
(m), 1386 (w), 1374 (w), 1335 (w), 1318 (w), 1300 (w), 1294 (w),
1263 (m), 1221 (w), 1159 (w), 1130 (m), 1095 (vs), 1067 (s), 1051
(s), 1015 (m), 991 (m), 974 (w), 939 (w), 930 (w), 897 (w), 866 (w),
841 (w), 802 (m, br), 764 (m), 734 (w), 723 (w), 646 (w), 621 (m),
584 (w); HRMS (ESI) calcd 313.1040 (M2+), found 313.1054. Anal.
Calcd for C26H28Cl2HgN2O8: C, 37.90; H, 3.42; N, 10.20. Found: C,
37.96; H, 3.23; N, 10.25.
tained in 3 mL quartz cuvettes with 1 cm path lengths (Starna) and
maintained at 25 °C by means of a circulating water bath.
Water Collection. Natural water samples from the Charles River
(Cambridge, MA), Onondaga Lake (New York state), and the harbor
in Newburyport, MA, were collected in August 2005 and stored in
polypropylene bottles. The water samples were passed through 0.2 µm
filters to remove particulate and insoluble organic matter prior to use.
Results and Discussion
X-ray Crystallographic Studies. Single crystals were mounted on
the tips of glass fibers coated with Paratone-N oil and cooled to -100
°C under a stream of N2 maintained by a KRYO-FLEX low-temperature
apparatus. Intensity data were collected on a Bruker APEX CCD
diffractometer with graphite-monochromated Mo KR radiation (λ )
0.71073 Å), controlled by a Pentium-based PC running the SMART
software package.49 Data collection and reduction protocols are
described elsewhere.50 Empirical absorption corrections were applied
by using the SADABS51 program, and the structures were solved by
direct methods using the SAINTPLUS52 and SHELXTL53 software
packages. The structures were checked for higher symmetry by using
the PLATON software package.54 All non-hydrogen atoms were located
and their positions refined with anisotropic thermal parameters by least-
squares cycles and Fourier syntheses. Unless otherwise noted, hydrogen
atoms were assigned to idealized positions and given thermal parameters
equivalent to either 1.5 (methyl hydrogen atoms) or 1.2 (all other
hydrogen atoms) times the thermal parameter of the carbon atom to
which they were attached. The structure of 6 includes one ordered
methanol and two ordered perchlorate anions. The O1 and N1 hydrogen
atoms of 6 were located from an electron density map. The structure
of 7 includes two ordered perchlorate groups.
Spectroscopic Materials and Methods. Millipore water (18.2 MΩ·
cm at 25 °C) obtained from a Milli-Q Biocel purifier outfitted with a
Quantum VX cartridge was used to prepare all aqueous solutions.
Piperazine-N,N′-bis(2-ethanesulfonic acid) (PIPES), N-(2-hydroxy-
ethyl)piperazine-N′-(2-ethanesulfonic acid) (HEPES), 2-(N-cyclo-
hexylamino)ethanesulfonic acid (CHES), and 4-cyclohexylamino-1-
butanesulfonic acid (CABS) buffers were purchased from Sigma and
used as received. Puratonic grade KCl was purchased from Calbiochem.
Mercury stock solutions (10 mM) were prepared from 99.998%
anhydrous HgCl2 (Aldrich) and water. DMSO stock solutions of 3 (2
mM, deep red-purple) and MS5 (1 mM, deep purple) were prepared,
partitioned into ∼40 µL aliquots, stored at -25 °C, and thawed in the
dark immediately before use. For measurements conducted in buffered
solutions, the buffer concentration was 50 mM and, with the exception
of the anion variation experiments, 100 mM KCl. The pH titrations
were conducted in the presence of 100 mM KCl. Quantum yields were
measured relative to fluorescein in 0.1 N NaOH (Φ ) 0.95).55
Experimental details for all spectroscopic measurements are available
elsewhere.34,43 All measurements were repeated a minimum of three
times, and the resulting averages are reported.
Design and Synthesis of MS5. Several considerations
influenced the design of sensor MS5. Our first objective was
to employ a fluorophore with relatively low-energy emission
relative to fluorescein and other commonly employed chro-
mophores. Metal ion sensors based on seminaphthofluorescein
3 (Figure 1) offer maximum emission at >600 nm in aqueous
solution, a feature beneficial for applications requiring multicolor
detection and for biological use where lower energy excitation
is preferred. A second objective was to obtain a sensor with
high selectivity for Hg(II) compared to other metal ions. Ligand
1 was previously employed in the synthesis of MS1 (Figure 1)
and confers better selectivity for Hg(II) than metal-coordinating
groups with fewer sulfur donor atoms. Scheme 1 shows the final
transformation in the nine-step assembly of MS5. Combination
of aniline 1 and aldehyde 2 in a mixed CHCl3/MeOH solvent
followed by imine reduction using NaB(OAc)3H afforded sensor
MS5 in 32% yield as a dark purple powder following purifica-
tion by preparative TLC on silica gel (20:1 CHCl3/MeOH).
We also prepared a salicylaldehyde-based analogue of MS5,
L, shown in Scheme 1, for use in X-ray crystallographic
modeling studies. Ligand L was obtained in 65% yield as an
off-white solid following Schiff base condensation of salicyl-
aldehyde with aniline 1 and reduction with NaB(OAc)3H in
DCE.
Spectroscopic Properties and Hg(II) Response of MS5.
Like fluorescein, seminaphthofluorescein (SNAFL)56 chro-
mophores exhibit pH-sensitive absorption and emission spectra.
The pH 7-9 range is particularly important because seminaph-
thofluorescein platform 3 (Figure 1) undergoes a fluorescence
change corresponding to a pKa of ∼8 in aqueous solution
containing 100 mM KCl (Figure S1, Supporting Information).
Decreasing the pH in this range results in fluorescence quench-
ing (Figure S1). This behavior is similar to the fluorescence
quenching observed upon protonation of fluorescein, namely,
fluorescence quenching (pKa ∼6.4). The corresponding changes
in the optical absorption spectra, depicted in Figure S2, are also
analogous to those observed upon fluorescein protonation to
form the fluorescein monoanion.57 From these comparisons, we
conclude that the seminaphthofluorescein monoanion is the
predominant species at pH 7, whereas the more emissive dianion
is present at higher pH.
Fluorescence spectra were collected by using a Photon Technology
International (Lawrenceville, NJ) Quanta Master 4L-format scanning
spectrofluorimeter equipped with an LPS-220B 75 W xenon lamp and
power supply, an A-1010B lamp housing with integrated igniter, a
switchable 814 photon-counting/analog PMT detector, and an MD-
5020 motor driver. Optical absorption spectra were collected on a Cary
1E double-beam scanning spectrophotometer. All samples were con-
We next describe the spectroscopic characteristics and
Hg(II) response of MS5 at pH g 7. Table 1 lists the absorp-
tion and emission properties of MS5 in the absence and in the
presence of excess Hg(II). The changes in optical absorp-
tion of free MS5 as a function of pH are comparable to those
of 3. There are two local maxima (500, 528 nm) of similar
intensity at pH 7 and one red-shifted local maximum (552 nm)
of much greater intensity at pH 9 (Figure S3). Similar pH-
(49) SMART: Software for the CCD Detector System, version 5.626; Bruker
AXS: Madison, WI, 2000.
(50) Kuzelka, J.; Mukhopadhyay, S.; Spingler, B.; Lippard, S. J. Inorg. Chem.
2004, 43, 1751-1761.
(51) Sheldrick, G. M. SADABS: Area-Detector Absorption Correction; Uni-
versity of Gottingen: Gottingen, Germany, 2001.
(52) SAINTPLUS: Software for the CCD Detector System, version 5.01; Bruker
AXS: Madison, WI, 1998.
(53) SHELXTL: Program Library for Structure Solution and Molecular
Graphics, version 6.1; Bruker AXS: Madison, WI, 2001.
(56) Whitaker, J. E.; Haugland, R. P.; Prendergast, F. G. Anal. Biochem. 1991,
(54) Spek, A. PLATON,
A
Multipurpose Crystallographic Tool; Utrecht
194, 330-344.
University: Utrecht, The Netherlands, 2000.
(57) Sjo¨back, R.; Nygren, J.; Kubista, M. Spectrochim. Acta Part A 1995, 51,
L7-L21.
(55) Brannon, J. H.; Madge, D. J. Phys. Chem. 1978, 82, 705-709.
9
J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007 5913

A R T I C L E S
Nolan and Lippard
Table 1. Spectroscopic Properties of MS5a
(Figure 1).48,58 We attribute the effect to protonation of the
aniline unit, which disrupts photoinduced electron transfer (PET)
quenching. The quantum yields for apo MS5 were determined
in the pH 7-9 range and are sensitive to pH changes in this
interval. At pH 7, Φfree is 0.018, and its magnitude decreases
to ca. 0.01 at pH g 8.
absorption
emission
λ
(nm); Φb
1
λ
(nm); ꢀ × 103 (M-1 cm-
)
pH
unbound
Hg(II)
unbound
Hg(II)
7
500; 4.5
528; 4.5
531; 12.2
552; 24.2
492; 7.0
533, 7.8
541; 16.0
547; 27.2
536, 611; 0.018
524, 612; 0.032
8
9
532, 612; 0.009
522, 620; 0.010
546, 624; 0.064
534, 624; 0.086
The fluorescence of MS5 is enhanced upon Hg(II) coordina-
tion at pH g 7. Values for ΦHg are listed in Table 1, and
representative emission spectra obtained for MS5 before and
after Hg(II) addition at pH 7, 8, and 9 are given in Figure 3. At
pH 7, an ∼4-fold fluorescence enhancement occurs following
Hg(II) coordination (ΦHg ) 0.032). At pH 8 (ΦHg ) 0.064)
and 9 (ΦHg ) 0.086), the increase in integrated emission is more
dramatic, with an ∼9- and ∼11-fold turn-on, respectively.
Interestingly, the nature of the fluorescence enhancement
depends on pH. Significant increases in both the 524 and 624
nm emission bands contribute to the fluorescence turn-on at
pH e 8. At higher pH, Hg(II) coordination gives rise to only
negligible changes in the 524 nm band, with dramatic increases
at 624 nm. This property affords single-excitation, dual-emission
ratiometric detection of Hg(II) at pH > 8 through comparison
of the ratio of the 524 and 624 nm (λ624/λ524) bands before and
after Hg(II) addition (λex ) 499 nm). For instance, an ∼13-
fold ratio change occurs at pH 9. This feature may facilitate
quantification in applied work.
a All measurements were made in the presence of 100 mM KCl and
with 50 mM PIPES (pH 7), 50 mM HEPES (pH 8), or 50 mM CHES (pH
9) buffer. b Fluorescein (Φ ) 0.95 in 0.1 N NaOH, ref 55) was used as the
standard for the quantum yield measurements.
The data included in Table 1 indicate that the greater dynamic
range in the fluorescence response resulting from Hg(II)
coordination at higher pH primarily stems from brighter Hg(II)
complexes rather than reduced background emission. This
observation contrasts the well-documented behavior of many
proton-sensitive PET-based sensors, which display significantly
reduced emission in alkaline solution.59,60 Although the photo-
physical details of this phenomenon warrant further investiga-
tion, we propose that the underlying cause of the behavior stems
from the protonation state of the seminaphthofluorescein
platform and that the seminaphthofluoresein dianion is required
for brighter emission from the MS5:Hg(II) complex. At pH 7,
the seminaphthofluorescein is protonated and monoanionic,
hence less emissive. Coordination of MS5 to Hg(II) does not
promote deprotonation of this moiety (vide infra), and, although
fluorescence turn-on occurs, emission from the weakly fluo-
rescent monoanonic form of the seminaphthofluorescein results.
At pH > 8, the seminaphthofluorescein is predominantly in its
dianionic and most emissive form. The aniline unit effectively
quenches its emission, as evidenced by the Φfree values in Table
1, and Hg(II) coordination restores emission from the dianion.
Figure 2. Effect of pH on the emission of 5 µM MS5 (100 mM KCl, λex
) 499 nm, T ) 25 °C). (A) The pH was decreased from ∼12 to ∼9 in
increments of ∼0.25. (B) The pH was decreased from ∼9 to ∼6 in
increments of ∼0.25. See Figure S4 for data taken at pH < 6.
As observed for MS1-4, the fluorescence response of MS5
to Hg(II) is sensitive to the presence of chloride ion. Figure 4
shows the fluorescence response of MS5 to Hg(II) in the
presence and in the absence of chloride ion at pH 8. In 50 mM
HEPES, an ∼3-fold fluorescence enhancement occurs upon
Hg(II) coordination, and subsequent introduction of 50 mM
chloride ion immediately evokes full turn-on. Substitution of
KCl with KX (X ) NO3-, OAc-, F-) reveals that this effect is
chloride-ion-specific, although acetate also gives some emission
dependent trends are observed for the MS5:Hg(II) complex
(Table 1, Figure S3).
The emission spectrum of free MS5 shows some variations
with pH that differ from those of 3. At pH g 9, MS5 exhibits
two emission bands centered at ca. 524 and 624 nm of similar
intensity (Figure 2). Decreasing the pH to ∼6 enhances the 524
nm band and reduces the 624 nm band with an isoemissive point
at 600 nm. Further acidification results in fluorescence quench-
ing. An ∼50% increase in integrated emission occurs as the
pH is lowered from ∼9 to ∼6, which contrasts with the
fluorescence decrease observed for 3 with increasing acidity.
This behavior is similar to that observed for other metal ion
sensors containing aniline-derivatized ligands, including fluo-
rescein-based MS134 and seminaphthofluorescein-based ZNP1
(58) Nolan, E. M.; Lippard, S. J., unpublished results.
(59) de Silva, A. P.; Gunaratne, H. Q. N.; Gunnlaugsson, T.; Huxley, A. J. M.;
McCoy, C. P.; Rademacher, J. T.; Rice, T. E. Chem. ReV. 1997, 97, 1515-
1566.
(60) Callan, J. F.; de Silva, A. P.; Magri, D. C. Tetrahedron 2005, 61, 8551-
8588.
9
5914 J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007

Mercury Sensing in Water with a Red-Emitting Probe
A R T I C L E S
Figure 3. Fluorescence response of MS5 to Hg(II) as a function of pH. The dashed lines represent the emission from 5 µM MS5 at pH 7 (50 mM PIPES,
100 mM KCl), pH 8 (50 mM HEPES, 100 mM KCl), and pH 9 (50 mM CHES, 100 mM KCl). The solid lines represent the emission change that occurs
immediately upon addition of 20 equiv of HgCl2. The samples were excited at 499 nm and T ) 25 °C.
refinement as well as selected bond lengths and angles,
respectively, for the L:Hg(II) complex 6. The asymmetric unit
contains the mercury complex, two perchlorate anions, and one
molecule of methanol. The N(1) and O(1) hydrogen atoms were
located from an electron density map. An ORTEP diagram of
6 (Figures 5 and S5) reveals the Hg(II) center to be five-
coordinate, with a distorted trigonal bipyramidal geometry. The
protonated phenol oxygen and tertiary amine nitrogen atoms
are in the axial positions, and the two thioether donors and the
protonated aniline nitrogen atom form the equatorial plane. The
Hg-S(1) and Hg-S(2) bond lengths are 2.5076(8) and
2.4854(8) Å, respectively (Table 2). These values are within
the general 2.47-2.73 Å range for Hg(II)-S(thioether) bonds
Figure 4. Effect of anions on the emission of the MS5:Hg(II) complex.
(A) The dashed line represents the emission from 5 µM MS5 in the absence
and indicate strong Hg-S interactions.61 The Hg-O(1) bond
length of 2.572(3) Å suggests a relatively weak interaction. The
Hg-N(2) distance, 2.436(3) Å, is slightly longer than that of
the analogous Hg-N bond in [Hg(beppa)(ClO4)]ClO4 (where
beppa ) N,N-bis(2-ethylthio)ethyl-N-[(6-pivaloylamido-2-py-
ridyl)methyl]amine), and the bond length of relevance is
2.412(6) Å.66 Overall, the structure of 6 is similar to that of
[(beppa)Hg(ClO4)]ClO4, although mercury in the latter is
described as six-coordinate because of a weak interaction with
a perchlorate ion (3.01 Å). The beppa ligand itself provides a
distorted trigonal bipyramidal geometry, where the tertiary amine
and amide oxygen atoms are in the axial positions and two
thioether donors and a pyridyl nitrogen atom lie in the equatorial
plane. The perchlorate groups of 6 are non-interacting, the
closest Hg-OClO3 interaction being 3.87 Å. Although the
model compound 6 suggests a five-coordinate geometry for the
Hg(II) complex of MS5, in aqueous solution a water molecule
(neutral pH) or hydroxide ion (basic pH) might be weakly
associated with, if not bound to, the Hg(II) center.
The EC50, or concentration of Hg(II) required to bring about
50% of the maximum fluorescence enhancement for a 1 µM
solution of the probe, was determined for MS5 at pH 7 (50
mM PIPES, 100 mM KCl). The maximum fluorescence
enhancement occurs in the presence of ∼10 equiv of Hg(II),
and the EC50 value is 910 nM (Figure S6). Despite the conserved
metal binding unit, this value is higher than that of MS1 (EC50
) 410 nM, pH 7). The difference may reflect different
protonation states of the fluorescein (MS1) and seminaphtho-
fluorescein (MS5) platforms at pH 7, with phenolate coordina-
tion to Hg(II) for MS1 and phenol coordination to MS5. When
a 0.5 µM solution of MS5 and 4 nm slit widths were employed,
of chloride ion (50 mM HEPES, pH 8). Addition of 20 equiv of Hg(II)
results in ∼3-fold fluorescence enhancement. Full emission is restored
immediately after addition of aqueous KCl (final concentration ) 100 mM).
(B) The effect of various anions in the buffer (50 mM HEPES, 100 mM
KX, pH 8): 1, no potassium salt; 2, KNO3; 3, KCl; 4, KOAc; 5, KF. F is
the fluorescence observed after Hg(II) addition, and Fo is the emission from
free MS5. The samples were excited at 499 nm and T ) 25 °C.
enhancement. The presence of chloride ion has no effect on
the emission of the free sensor. We have encountered such
chloride ion dependence for Hg(II) sensors with a variety of
ligand frameworks and have proposed that formation of a Hg-
Cl bond or strong ion-pairing influences the degree of turn-on
in these systems.34,35,43 Nevertheless, such chloride ion depen-
dence is not completely general. For instance, some xanthenone-
substituted Zinpyr sensors that contain di(2-picolyl)amine
ligands, such as ZP3 (Figure 1), give fluorescence enhancement
following Hg(II) coordination, and the degree of turn-on is
independent of chloride ion in the buffer.58 Fluo-5N, a sensor
sold by Invitrogen, gives fluorescence turn-on for Hg(II), and
the response is enhanced ∼2.5-fold by substituting chloride with
nitrate ion (50 mM PIPES, 100 mM KX, pH 7).58 These
observations indicate that buffer composition must be considered
when evaluating the response of fluorescent metal ion sensors.
Hg(II) Binding and Selectivity Studies. MS5 and MS1 share
the same N2S2O set of chelating donor atoms. A 1:1 binding
stoichiometry was previously observed in solution for Hg(II)
complexation to MS1. In this work, we prepared a salicylalde-
hyde-based model complex, L (Scheme 1), with the aim of
elucidating further details about the likely mode of Hg(II)
coordination to these sensors. X-ray-quality crystals were
obtained by vapor diffusion of Et2O into a methanolic solution
containing a 1:1 L/Hg(ClO4)2 mixture at room temperature.
Tables S1 and S2 contain crystallographic data from the
(61) Wright, J. G.; Natan, M. J.; MacDonnell, F. M.; Ralston, D. M.; O’Halloran,
T. V. Prog. Inorg. Chem. 1990, 38, 323-412.
9
J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007 5915

A R T I C L E S
Nolan and Lippard
Figure 5. ORTEP diagrams of the Hg(II) complexes 6 and 7. The ORTEP diagrams of 6 and 7 depict the 40% probability thermal ellipsoids for all
non-hydrogen atoms. The left and middle ORTEP diagrams omit the perchlorate counterions. The right-most ORTEP diagram of 7 illustrates the orientation
of the weakly associated perchlorate anion. The Hg(1)-O(1) distance is 2.715(5) Å.
Table 2. Selected Interatomic Distances (Å) and Angles (deg)a
[Hg(L)2](ClO4)2‚CH3OH (6)
Hg1-N1
2.334(3)
2.436(3)
2.5076(8)
2.4854(8)
2.572(3)
79.18(9)
114.59(7)
120.97(7)
N1-Hg1-N2
N2-Hg1-O1
N2-Hg1-S1
N2-Hg1-S2
O1-Hg1-S1
O1-Hg1-S2
S1-Hg1-S2
87.97(9)
166.63(8)
83.18(6)
83.35(6)
105.42(6)
100.05(6)
121.98(3)
Hg1-N2
Hg1-S1
Hg1-S2
Hg1-O1
N1-Hg1-O1
N1-Hg1-S1
N1-Hg1-S2
[Hg(TPEN)](ClO4)2 (7)
Hg1-N1
2.470(5)
2.434(5)
2.337(5)
2.572(5)
2.360(5)
2.357(5)
74.55(17)
72.97(18)
70.86(16)
123.49(17)
109.31(18)
N2-Hg1-N3
N2-Hg1-N4
N2-Hg1-N5
N2-Hg1-N6
N3-Hg1-N4
N3-Hg1-N5
N3-Hg1-N6
N4-Hg1-N5
N4-Hg1-N6
N5-Hg1-N6
131.58(17)
126.37(17)
70.53(18)
72.45(17)
73.57(17)
99.54(18)
153.11(18)
162.29(18)
81.85(17)
100.74(18)
Hg1-N2
Hg1-N3
Hg1-N4
Hg1-N5
Hg1-N6
N1-Hg1-N2
N1-Hg1-N3
N1-Hg1-N4
N1-Hg1-N5
N1-Hg1-N6
Figure 6. Reversibility of Hg(II) coordination to 5 µM MS5 by TPEN
addition (50 mM PIPES, 100 mM KCl, pH 7). The top set of solid lines
represent the fluorescence enhancement that occurs immediately after
addition of 10 equiv of Hg(II). The bottom set of solid lines represents the
emission from free MS5 and the emission decreases that occur immediately
after addition of 10 equiv of TPEN to a solution containing the MS5:
Hg(II) complex. Six cycles of on/off by Hg(II)/TPEN addition are depicted
in this plot, and the slight fluorescence rise observed in the bottom spectra
at ca. 530 nm results from TPEN emission (control data not shown). The
samples were excited at 499 nm and T ) 25 °C.
a Atoms are labeled as indicated in Figures 5, S5, and S7.
a lower detection limit of 50 nM Hg(II) was obtained at pH 7
(21% ( 4.5% fluorescence enhancement, mean ( SD for 16
trials).62
Binding of MS5 to Hg(II) is readily reversible. Addition of
TPEN to solutions of MS5 and Hg(II) caused an immediate
fluorescence decrease to within ∼20% of the baseline value
(Figure 6). This behavior is expected because TPEN has a
reported Kd value of ∼10-25 M for Hg(II) at 100 mM ionic
strength.63 Consecutive additions of Hg(II) and TPEN revealed
that this on/off behavior could be cycled a number of times
(Figure 6). Addition of excess KI(aq) also reverses the Hg(II)
binding (data not shown).43,64 Such reversibility and regeneration
are important for the fabrication of devices to sense mercuric
ion.
quality block crystals were grown from vapor diffusion of Et2O
into a methanolic solution of 1:1 TPEN/Hg(ClO4)2. ORTEP
diagrams of 7 are given in Figures 5 and S7; Tables S1 and 2
contain crystallographic data and select geometric parameters.
The Hg(II) center is seven-coordinate, including one weakly
bound perchlorate ion which has a Hg-O(1) distance of
2.715(5) Å. The stereochemistry of the Hg(II) center is that of
a C2V capped trigonal prism. The tertiary amine Hg-N(1) and
Hg-N(2) bond lengths are 2.470(5) and 2.434(5) Å, and the
bonds between the Hg(II) center and the pyridyl groups are
shorter, ca. 2.35 Å, and similar in length to that in [Hg(beppa)-
ClO4)]ClO4.66
No crystal structure determination of a TPEN:Hg(II) complex
has been published to date, and structures of TPEN:Zn(II)
complexes have only recently been reported.65 Colorless X-ray-
Figure 7 depicts the results of metal ion selectivity studies
for MS5 at pH 7. The selectivity is analogous to that observed
for MS1,34 with only Cu(II) interfering with the fluorescence
response to Hg(II). Variations in pH from 7 to 9 have no effect
(62) A 500 nM concentration of MS5 was employed because we were unable
to obtain reliable data with lower concentrations and 4 nm slit widths.
(63) Anderegg, G.; Hubmann, E.; Podder, N. G.; Wenk, F. HelV. Chim. Acta
1977, 60, 123-140.
(64) Coronado, E.; Ga´lan-Mascaro´s, J. R.; Mart´ı-Gastaldo, C.; Palomares, E.;
Durrant, J. R.; Vilar, R.; Gratzel, M.; Nazeeruddin, M. K. J. Am. Chem.
Soc. 2005, 127, 12351-12356.
(66) Makowska-Grzyska, M. M.; Doyle, K.; Allred, R. A.; Arif, A. M.; Bebout,
(65) Mikata, Y.; Wakamatsu, M.; Yano, S. Dalton Trans. 2005, 545-550.
D. C.; Berreau, L. M. Eur. J. Inorg. Chem. 2005, 822-827.
9
5916 J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007

Mercury Sensing in Water with a Red-Emitting Probe
A R T I C L E S
Figure 8. Fluorescence response to Hg(II) and Zn(II) from a mixture of 5
µM MS5 and 100 nM QZ2 at pH 7 (50 mM PIPES, 100 mM KCl). The
black dashed line represents emission from the MS5/QZ2 mixture. The red
line is the emission from the mixture immediately after addition of only 50
µM Hg(II). The green line is the emission from the mixture immediately
after addition of only 50 µM Zn(II). The blue line is the response in the
presence of both 50 µM Hg(II) and Zn(II). The samples were excited at
499 nm and T ) 25 °C.
Figure 7. Selectivity of MS5 for Hg(II) over other cations at pH 7 (50
mM PIPES, 100 mM KCl). (Top) Fluorescence response of 5 µM MS5
following addition of 20 equiv of the cation of interest: 1, LiCl; 2, NaCl;
3, RbCl; 4, MgCl2; 5, CaCl2; 6, SrCl2; 7, BaCl2; 8, MnCl2; 9,
Cr(CH3CO2)3; 10, Fe(NH4)2(SO4)2; 11, CoCl2; 12, NiCl2; 13, Cu(SO4)2;
14, ZnCl2; 15, CdCl2; 16, HgCl2; 17, Pb(NO3)2. (Bottom) The light gray
bars correspond to the bars in the top plot. The black bars indicate the
fluorescence change that occurs immediately following addition of 20 equiv
of Hg(II) to the solutions containing MS5 and the cation of interest. All
data (F) were normalized with respsect to emission of the free sensor (Fo).
The samples were excited at 499 nm, and the emission was integrated from
505 to 700 nm. T ) 25 °C. MS5 is also selective for Hg(II) in the presence
of millimolar concentrations of the alkali and alkaline earth metals (data
not shown).
on the observed selectivity, which is also independent of ionic
strength and the choice of anion in the buffer.
Figure 9. Response of 5 µM MS5 to 10 equiv of Hg(II) added to natural
water samples. (A) Emission spectra for MS5 in Charles River water before
(dashed line) and after (solid line) addition of 10 equiv of Hg(II). (B)
Fluorescence response of MS5 to Hg(II) in Charles River water (CRW),
Newburyport water (NPW), and Onondaga Lake water (OLW). The response
(F) is normalized to the emission of the free sensor (Fo). The samples were
excited at 499 nm and T ) 25 °C.
Simultaneous Detection of Hg(II) and Zn(II). Multi-analyte
sensing is important for simultaneously determining the presence
of various components in a mixture.67 We therefore determined
whether MS5 could perform in combination with a Hg(II)-
insensitive fluoroionophore to detect multiple metal ions in
aqueous solution. We chose a fluorescein-based Zn(II) sensor,
QZ2 (Figure 1), which has Hg(II)-insensitive emission and
affords fluorescence enhancement for Zn(II) in the presence of
Hg(II).68 Figure 8 depicts the response of an MS5/QZ2 mixture
in the presence of Hg(II), Zn(II), and both Hg(II) and Zn(II) at
pH 7. Independent fluorescence changes of MS5 and QZ2 are
observed in the presence of only Hg(II) or Zn(II), respectively.
When a mixture of Hg(II) and Zn(II) is added, both sensors
respond independently to their respective analytes. The rise
centered at 624 nm indicates the presence of Hg(II), and that
centered at 520 nm signals Zn(II). This experiment demonstrates
that MS5 can be employed in conjunction with a sensor emitting
at shorter wavelengths for multi-analyte detection.
first step toward the former objective, we tested the ability of
MS5 to respond to Hg(II) in natural water. Samples were
collected from three sources, including the Charles River
(Cambridge, MA), and representative data are given in Figure
9. In each case, MS5 shows ∼4.5-fold or greater fluorescence
turn-on in samples spiked with Hg(II). This result indicates that
MS5 can detect Hg(II) in solutions with significantly more
complex composition than laboratory buffer and suggests that
its sensitivity to chloride ion will not be a limiting factor for
use in environmental monitoring.
Summary and Perspective
This work describes the synthesis and characterization of
mercury sensor MS5. This sensor was designed to afford low-
energy emission, both turn-on and ratiometric fluorescence
responses to Hg(II), and a high degree of Hg(II) ion selectivity.
The solution studies show that pH is a key factor for generating
a ratiometric response to Hg(II). We propose that this behavior
reflects the requirement of the seminaphthofluorescein dianion
for significant enhancement of the 624 nm band, although more
MS5 Performance in Natural Water Samples. Small
molecule fluorescence detectors for Hg(II) have the potential
for use in the field. Such an application presents a unique set
of challenges and requires detailed studies of sensor performance
in the environmental milieu and method/device design. As a
(67) Walt, D. R. Curr. Opin. Chem. Biol. 2002, 6, 689-695.
(68) Nolan, E. M.; Jaworski, J.; Okamoto, K.-I.; Hayashi, Y.; Sheng, M.; Lippard,
S. J. J. Am. Chem. Soc. 2005, 127, 16812-16823.
9
J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007 5917

A R T I C L E S
Nolan and Lippard
detailed physical characterization is needed to address this
notion. It became evident during our investigation that the
fundamental chemistry of seminaphthofluorescein is relatively
unexplored.69 Detailed investigations of its protonation and
tautomeric equilibria and fluorescence behavior in a number of
different solvent systems will be especially helpful in both
creating new red-emitting chromophores with enhanced bright-
ness at neutral pH and guiding future sensor design. MS5 gives
selective fluorescent enhancement for Hg(II), binds Hg(II)
rapidly and reversibility, and has a lower detection limit of 50
nM at pH 7. Preliminary studies in natural water samples
indicate that MS5 offers immediate Hg(II) detection in complex
media. This observation points to its potential utility in the field.
Acknowledgment. This work was supported by Grant
GM65519 from the National Institute of General Medical
Sciences. E.M.N. thanks the Whitaker Health Science Fund for
a graduate fellowship, Ms. C. W. Nolan for collecting water
samples, Drs. D. Song and R. P. Feazell for assistance with the
crystal structure determinations, and Prof. C. Fahrni for a helpful
discussion.
Supporting Information Available: Summary of X-ray
1
crystallographic data, Figures S1-S7, and representative H
NMR spectra (PDF); X-ray crystallographic files for 6 and 7
(CIF). This material is available free of charge via the Internet
at http://pubs.acs.org.
(69) A SciFinder search yielded only 39 hits and only 116 for “seminaphtho-
fluorescein” and only 115 for “SNAFL”, compared to over 60 000 hits
generated by “fluorescein” in November 2006.
JA068879R
9
5918 J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007